On dynamic negotiation strategy for concurrent negotiation over distinct objects

This paper addresses the problem of generating counteroffers by an agent negotiating with multiple opponents concurrently over multiple distinct objects characterized by multiple issues (attributes.) Most previous works address negotiation strategies for simpler situation where an agent negotiates with multiple opponents for the purpose of securing an agreement over a single object with either a single or multiple negotiation issues. We propose a novel dynamic negotiation strategy that works in a more complicated negotiation scenario. The strategy involves adaptation of both, the initially generated counteroffers and the issues counteroffers' weights matrix during negotiation. The proposed dynamic strategy takes into consideration the behaviors of the current opponents in terms of their recent concessions to fine-tune the negotiation strategy of the agent in real time. The initial experimental results show that the proposed mechanism is effective in terms of both, the agreement and utility rates when compared with a static strategy.